Unpacking Multifamily Supply Risks and Demand Booms

Analyzing Multifamily Construction Risk Alongside Resurgent Demand 

As multifamily occupancies retreated through 2023, much attention was given to the looming wave of new construction. With more than 1 million units under construction at the peak in July of 2023—the most the U.S. has ever seen—the focus was understandable. Over the past year, the market experienced some of the strongest demand for apartments on record, second only to the post-pandemic surge. Now, with the construction peak in the rearview mirror or soon to be, the key questions are: What does this mean for multifamily moving forward? And which markets are ahead or behind the curve? 

Where is multifamily demand coming from? 

First, we need to establish why the market has experienced such strong absorption.  Across a number of previous reports External Link, we detailed the growing unaffordability of the single-family market. At the time, this wasn’t reflected in U.S. Census Bureau data—homeownership rates were rising even as affordability collapsed.  However, since mid-2023, the data converged with the situation on the ground: the overall homeownership rate has declined 30 basis points (bps) over the last year. This trend becomes clearer when indexing the homeownership rate by age, as homeownership typically increases with age, magnifying the observed decline.  

The declines in homeownership have been almost exclusively among people under 35 and those aged 35-44. These age groups already have the lowest homeownership rates, so their declines significantly contribute to rising renter household formation.  

Other long-term demographic shifts are also driving renter household growth. The biggest contributors to household formation are all three age groups under 44, and this is further compounded by the decline in the homeownership rate. At the same time, fewer 25-34-year-olds are living with their parents. While the decline leveled off in 2023, it remains 2% below its peak. If the decline continues and returns to its historical average, it would lead to an additional 1.2 million new households—most likely renters. If the trend reaches pre-2010 averages, that figure would jump to 1.8 million new households, including the 1.2 million previously mentioned.  

Household sizes have consistently declined after remaining mostly consistent from 2010 to 2020. The figures show relatively minor moves, but small shifts in average household size can have a meaningful impact on household formation, especially given the millions of Americans who rent. Additionally, the Federal Reserve Bank of New York’s consumer survey shows that the percentage of people expecting to move within the next 12 months External Link has risen throughout 2024, increasing from 13.6% to over 18%, the highest level since the pandemic.  

Consistent with overall population growth, most renter households are being formed in the Sun Belt. Of the top 25 markets for apartment demand over the trailing 12 months (TTM), 11 are in the Southern U.S. Dallas leads the nation with more than 16,000 units absorbed, while Houston, Austin and Atlanta each had more than 10,000 net move-ins over the same time frame. 

What about construction? 

Now that the supply wave is cresting, as detailed in the 2024 Q2 U.S. Multifamily MarketBeat, the risk of oversupply is dwindling. However, this trend varies by market. While cities like Miami, Charlotte and Austin still have some of the highest levels of supply in the nation, the overall pace of construction has decreased dramatically from a year ago.  

Nearly all markets—except Orange County, San Francisco, Pittsburgh and Richmond— have experienced a reduction in their pipelines shrink over the past year, with significant declines in some of the epicenters of development post-pandemic. Austin and Nashville posted the sharpest reductions in construction relative to their inventories. Austin’s construction fell from nearly 19% of its inventory last year to about 11% today, while Nashville’s fell from nearly 15% to just under 9%. While these cities have some of the most construction in the nation, significantly more product has delivered than has been started in most markets nationwide.  

Examining construction starts offers insight into which markets may see sharper recoveries. The graph below compares demand over the last year (the Y axis) alongside construction starts relative to pre-pandemic norms. While nearly all markets have new multifamily starts below their pre-pandemic averages, analyzing this data in relation to demand offers a hint at what the recovery curve will look like across U.S. metros.  

Where does that leave the market? 

Last year, in our report Contextualizing Development Risk, we introduced the concept we called “years of construction.” It borrows from the concept of “months of supply on market” from the for-sale residential space, which measures how many months it would take to sell all current listings at the current sales pace.  We applied this idea to estimate how many years it would take to absorb all new multifamily construction, given historical levels of demand.  

In the wake of significant disruption from the pandemic and subsequent migration surges, it became difficult to pinpoint where demand would head as the disruption waned. As such, our prior analysis used pre-pandemic demand figures as the denominator while also offering additional scenarios. With demand surging over the past year, it behooves an updated analysis, using more recent demand metrics.  

With construction pipelines dwindling over the last year, the balance of risk has shifted. Last year, six markets had more than five years’ worth of supply coming to market. Today, no markets have more than five years of construction left in their pipelines.  

However, this analysis has some limitations. Many of the nation’s tightest markets, like Southern California and New York, have experienced limited demand due to very few available units, despite having minimal new construction. As such, the previous chart showing percentage of inventory underway may better reflect the relative risk of these markets than the “years of supply” metric. As a result, markets with 3% or less of their inventory underway were excluded from the following charts.  

With about 8,000 units under construction and an absorption pace of about 2,200 units over the last year, both the Inland Empire and San Jose lead the nation in relative riskiness, with slightly more than three and a half years of construction remaining. It isn’t to say that either market is at risk of imminent oversupply conditions – indeed, the pipeline will empty over the next two to three years, but demand will need to accelerate as these units deliver to prevent an overhang of new supply.  

The “supply side” of the equation is incomplete in today’s environment, as many markets already face oversupply conditions. Once units deliver, the properties fall out of the analysis. To accommodate for both imperfections, we can consider “excess vacancy,” which measures the number of excess units that need to be leased to bring markets back to pre-pandemic (2019) vacancy levels. 

Here, we see a divergence between the two leading markets on the prior chart. While Inland Empire’s vacancy is above its 2019 mark, San Jose’s is below, showing relatively divergent riskiness between the two California markets. Conversely, markets like Austin and San Antonio shift up the risk curve as vacancies have ticked higher on the back of heavy deliveries. For these markets, the combination of elevated vacancies and deliveries adds an additional year for those markets to recover, assuming demand holds consistent. It’s notable that nearly every market on this year’s list has shifted down from a year ago, shortening the timeline as a result of strong demand formation.  

While some markets still have sizable construction pipelines, the uptick in demand suggests that the worst may be over for most markets. New construction starts have come in significantly across most markets, and assuming the economy avoids a recession while demand remains consistent, the degree of oversupply is expected to decrease quickly. This will likely lead to improvement in fundamentals and property values, even in a higher interest rate environment, according to our base case outlook.